Hierarchical fuzzy clustering based on self-organising networks

A fast and computationally efficient fuzzy clustering approach is presented. In this approach, fuzzy clustering is implemented in two hierarchical phases: subclusters generation by a self-organising network and fuzzy classification via a fuzzy competitive clustering network associated with a fuzzy c-means algorithm. Owing to the hierarchical network, the computation complexity of fuzzy clustering is reduced drastically and the clustering performance is enhanced as well. The simulation results show that the proposed method has a much higher computing efficiency and better classification performance compared to standard fuzzy c-means clustering.